Membrane dynamics of human red blood cells
parasitized by Plasmodium falciparum
YONGKEUN PARK1+, MONICA DIEZ-SILVA2+, GABRIEL POPESCU1, GEORGE
LYKORAFITIS2, WONSHIK CHOI1, MICHAEL S. FELD1 and SUBRA SURESH3*
1
G. R. Harrison Spectroscopy Laboratory, Massachusetts Institute of Technology, Cambridge,
Massachusetts 02139, USA
2
Department of Materials Science and Engineering, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139, USA
3
School of Engineering and Harvard-MIT Division of Health Science and Technology, Massachusetts
Institute of Technology, Cambridge, Massachusetts 02139, USA
+These authors contributed equally to this work.
*
e-mail: ssuresh@mit.edu
Abstract
Cell membrane dynamics, which is influenced by metabolic activity, stresses, and
cytoskeltal defects, is a possible indicator of disease. Here we report the first
experimental evidence of connections between changes in membrane fluctuation
amplitudes and onset of disease state. Through diffraction phase microscopy (DPM), we
quantify membrane fluctuations of human red blood cell (RBC) parasitized by
Plasmodium falciparum merozoite at 37 ºC and 41ºC relevant to malaria fever episode
cycles. The membrane fluctuations measured at spatial and time resolutions of nm and
ms, respectively, cover the entire range of intra-erythocytic developmental stages of the
parasite. DPM results are also used to calculate the membrane shear modulus as a
function of parasite development at both 37 ºC and 41ºC. Our results indicate that
pararsite modification of host RBC membrane and internal structure cause progressive
changes in RBC membrane fluctuations and elastic deformability properties. Such
biophysical modifications become more pronounced in later stage parasite development
and at febrile temperature. These findings offer new mechanistic insights into cell
biophysical property changes of RBCs parasitized by P. falciparum and also provide
possible new avenues for identifying human diseases through cell membrane
dynamics.Cell membrane dynamics is widely recognized to be strongly influenced by the
onset and progression of human diseases. Studies of human red blood cells (RBCs) reveal
1
that amplitudes of the fluctuations in the phospholipid bilayer and attached spectrin
cytoskeleton are influenced strongly by cytoskeletal defects, mechanical stresses, and
actinspectrin dissociations arising from metabolic activity linked to adenosine 5triphosphate (ATP) concentration. Changes resulting from the transfer of proteins from
foreign microorganisms that invade the RBC, such as the malaria-inducing parasite P.
falciparum, to specific binding sites in the spectrin cytoskeletal network can also cause
significant alterations to RBC membrane dynamics. These fluctuation responses provide
insights into possible mechanistic pathways in the pathogenesis of malaria, as the parasite
alters the biophysical and biomechanical properties of RBC during its intra-erythrocyte
stage that lasts up to 48 h during the so-called asexual stage of malaria. Despite the broad
realization that membrane fluctuations provide information on critical interactions among
subcellular structure, biochemical links between the cell interior and the external
environment, and mechanical stresses on the cell, systematic and comprehensive
controlled experiments of cell membrane dynamics, especially over the physiologically
relevant temperature range, have not been performed in the context of specific human
diseases. We present experimental results of membrane fluctuations in human RBCs
parasitized by P. falciparum over the full range of intra-erythrocyte stages at both normal
physiological (37ºC) and febrile (41ºC) temperatures. We show that parasite interactions
with the host RBC cause strongly temperature-dependent and stage-specific alterations to
membrane dynamics.
We employed DPM to quantify the progressive alterations to the mechanical
response of an RBC following the invasion of the malaria-inducing P. falciparum. This
technique provides quantitative information about the membrane thermal fluctuations
which develop at the nanometer and millisecond scales. DPM uses interferometry to
extract the optical path-length shift produced at each point across the cell. This
information is translated into the cell thickness profile h( x, y ) , by taking into account the
optical homogeneity of the internal cell composition, i.e. refractive index. We employed
the tomographic phase microscopy (TPM) in order to retrieve the averaged refractive
index over the cytoplasm of P. falciparum infected RBCs as well as healthy ones. (see
Method section). Recently, this approach has been applied to study healthy RBCs under
both static and dynamic conditions. Preparation of RBCs and P. falciparum cultures for
2
DPM and TPM experiments are described in the Methods section.
Figs. 1a-d show DPM topographic images of healthy and infected RBCs at various
stages of parasite development (see also supplemental movie S3). As the parasite matures
within the host RBC, the normal discocyte shape is lost. Parasite modifications to the
membrane structure of infected RBCs are implicated in the observed morphological
changes (REF?). Assuming an elastic restoring force associated with the membrane
displacements, we extract an effective spring constant ke at each point on the cell
2
T
/
2

k
(
x
,)
y
h
(
x
,)/
y
2
B
e
membrane using k
, where kB is the Boltzmann constant, T
is the absolute temperature, and
h(x, y)2
is the mean-squared displacement.
Representative ke maps of parasite infected cells at the indicated stages are shown in Figs.
1e-h. These results show that membrane stiffness increases progressively with parasite
development. In particular, the spatially averaged effective stiffness, ke (x, y) , at the
schizont stage is a factor of five to ten higher than healthy RBCs. These trends are fully
consistent with those inferred from large-deformation stretching with optical tweezers of
Plasmodium-infected RBCs over all stages of parasite maturation. A histogram showing
membrane displacements based on DPM for all parasite stagesis given in Fig. 2a. These
results indicate that the stage of parasite development directly correlates with the
amplitude of membrane fluctuations.
RBC deformability is significantly dependent on membrane stiffness. Our DPM
experiments provide information to quantitatively measure determine the membrane
shear modulus. We measure the shear modulus to be 5.7 ± 0.9 μN/m for healthy RBCs
(Fig.2B), which is consistent with earlier works based on micropipette asipration and
optical tweezers (REFS?). The increase in shear modulus for ring (μ = 17.9 ± 5.7
μN/m)and trophozoite stages (μ = 29.8 ± 11.3 μN/m) is consistent with previous studies
(REFS?). The schizont stage has μ = 106.5 ± 36.5 μN/m.
In addition to tests conducted at physiological temperature (37oC), we measured
fluctuations of healthy and infected RBC membranes at 41oC to study the effect of febrile
temperature, commonly experienced during a malaria attack, on RBC membrane
3
dynamics. Because many previous studies on P. falciparum infected RBC mechanical
properties using other techniques were conducted at room temperature, we also present
DPM results from tests carried out under ambient temperature. The results are
summarized in Fig. 3. From room to physiological temperature, membrane fluctuations of
healthy RBC increase for all test conditions (Fig. 3), which is attributed to the increase in
available thermal energy. For healthy RBCs, membrane fluctuations increase consistently
from room temperature to physiological, and febrile temperature. Interestingly, there is
significant increase in membrane fluctuations (53% in FWHM value) from physiological
to febrile temperature (p < 0.01), which is only a 7.5% increase in absolute temperature..
Evidently, such enhancement in the fluctuation cannot be explained only by simple
equilibrium thermodynamics, i.e. the increase in the Boltzmann factor kBT. This suggests
that the RBC membrane undergoes structural changes that affect viscoelastic properties.
One possible explanation is that a transitional structural change to spectrin molecules
occurs between physiological and febrile temperature (REF?), resulting in a dramatically
changed cytoskeleton network organization.. Both α- and β- spectrin molecules have a
significant structural transition near 40ºC. It has previously been shown with the
micropippette aspiration technique that RBC membrane elastic shear modulus decreases
by about 20% when temperature increases from 25 ºC to 41ºC(REF?). From
physiological to febrile temperature membrane fluctuations decrease significantly (p <
0.01) for all infected test conditions (Fig. 3.), with trophozoite and schizont stages
showing the strongest temperature-dependence in this range. Based on comparison with
results from healthy cells, we speculate that intra-erythrocytic parasite development
causes an opposite effect. Parasite exported proteins that target the RBC membrane can
alter spectrin folding transitions involved in in stabilizing the erythrocyte cytoskeleton
(REF?). Expression of P. falciparum genes is known to be affected by exposure to 41ºC
(REF?). Under febrile conditions, increased levels of parasite exported proteins could
contribute to large changes in membrane fluctuations observed by DPM between
physiological and febrile temperature. One example is the ring-infected erythrocyte
surface antigen (RESA). When RESA binds spectrin, the dimer-tetramer ratio shifts in
favor of the tetramer, which increases membrane stability (REF?). As a result, RESA
enhances RBC resistance to mechanical and thermal stress. Moreover elevated thermal
4
stability conferred by RESA plays a protective role for infected-RBCs against damage at
febrile temperature. The development of adhesion properties at the trophozoite and
schizont stages (REF?) may also contribute to the significant decrease in fluctuation
observed by DPM at these stages, especially at febrile temperature. For the same
temperature, our results indicate that fluctuation amplitude progressivelydecreases as
parasite develops from the ring to the schizont stage.
In summary, we present the first systematic measurements of membrane fluctuations
associated with Plasmodium-infected RBCs at all stages of parasite development and for
physiological and febrile temperatures. Our approach to studying the parasitized RBCs
uniquely combines optical interferometry, biophysics, continuum mechanics, and
microbiology. Membrane fluctuation results can also be used to extract elastic
deformability properties of RBCs infected by P. falciparum. Compared to other
techniques for assessing RBC mechanical properties such as electric field deformation,
pipette aspiration, optical tweezers, and magnetic bead excition, the method presented
here has distinct advantages of being spatially-resolved and non-contact. We envision
that this methodology can be used as a tool for studying the pathogenesis of malaria and
monitoring the effects of various drugs aimed at reversing the mechanical degradation of
RBC membranes. In addition, diffraction phase microscopy can provide the necessary
data for testing theoretical models and computational simulations pertinent to a broader
range of cell biology problems, way beyond the pathogenesis of malaria.
METHODS
Preparation of red blood cells and parasite culture
Plasmodium falciparum (3D7) parasites were maintained in leukocyte-free human O+
erythrocytes (Research Blood Components, USA) and stored at 4°C for no longer than
two weeks under an atmosphere of 3% O2, 5% CO2 and 92% N2 in RPMI 1640 medium
(Gibco Life Technologies) supplemented with 25 mM HEPES (Sigma), 200 mM
hypoxanthine (Sigma), 0.209% NaHCO3 (Sigma) and 0.25% albumax I (Gibco Life
Technologies). Cultures were synchronized successively by concentration of mature
schizonts using plasmagel flotation and sorbitol lysis two hours after merozoite invasion
to remove residual schizonts (REFS?).
5
Healthy RBC control samples and P. falciparum infected RBC samples were diluted
in PBS (phosphate buffer solution) to approximately 106 RBC/ml prior to membrane
fluctuation experiments. Measurements were performed at 1420 h (ring stage), 2036 h
(trophozoite stage), and 3648 h (schizont stage) following merozoite invasion. 30 cells
were measured for each stage of parasite development and for each test temperature.
Healthy and infected RBCs at specific stages of parasite development were placed in
glass wells and imaged by DPM. When RBCs are positioned on the glass substrate, the
membrane in contact with the substrate can become attached. However, the effect of
membrane attachment to the substrate was found to negligibly affect membrane
fluctuation measurement. Typically, samples parasitemia was around 5%. To identify
infected cells from non-infected cells, we use DAPI staining to identify infected RBCs.
Prior to DPM dynamic measurements, we recorded epi-fluorescence images (for details
see referece (REF?)). (See Supplemental Material for full details of DAPI staining).
DPM imaging of RBCs
An Ar++ laser (=514 nm) was used as an illumination source for an Olympus IX71
inverted microscope. The microscope was equipped with a 40X objective (0.65 NA),
which grants a diffraction-limitted transverse resolution of 400 nm. With the additional
relay optics used outside the microscope, the overall magnification of the system was
approximately 200. DPM employs the principle of laser interferometry in a common path
geometry and thus provides full-field quantitative phase images of RBCs with
unprecedented optical path-length stability. The instantaneous cell thickness map is

(
x
,,
y
t
)


n
x
,,
y
t
)
/2
(
obtained as h
, with  the quantitative phase image measured
by DPM. The refractive index contrast n between the RBC and surrounding PBS is
mainly caused by the presence of hemoglobin (Hb), which is optically homogeneous in
cytosol. P. falciparum infected RBCs, however, are not optically homogeneous, caused
by many factors that affect the refractive index : the parasite itself occupies a large
volume in the cytosol of the host RBC; loss of Hb due to conversionto hemozoin crystal
by the parasite; and, the addition of various parasite exported proteins into the host RBC
cytosol. We employed TPM to retrieve the three-dimensional distribution of refractive
6
index for all the stages of P. falciparum infected RBCs as well as healthy RBCs24. The
refractive index contrast, n, in the cytosol, excluding the parasite and hemozoin, was
found to be is 0.059±0.002, 0.058±0.006, 0.056±0.005, and 0.044±0.002 for healthy `,
ring-stage, trophozoite, and schizont stage RBCs. The DPM optical path-length stability
is 2.4 mrad, which corresponds to a membrane displacement of 3.3 nm. TPM provides
three dimensional distribution of refractive index with a sensitivity of 0.001 rad.
Temperature control
The microscope was equipped with a temperature controller (TC-202A, Warner
Instruments) that uses a thermistor to set the sample temperature within ±0.2°C. The
sample well is placed in contact with the controller chamberto maintain thermal
equilibrium between the two systems. Equilibrium is attained within 3-4 minutes and
DPM tests were commenced after 10 minutes.
Membrane displacement analysis
The instantaneous cell displacement, h(x, y, t) map was obtained by subtracting
the time-averaged cell shape from each thickness map in the series. The effective spring
2
(
x
,y
)

k
T
/
h
(
x
,y
)
e
B
constant map of the cell is obtained as k
.
The shear modulus μ can be obtained using the Fourier-transformed Hamilotian
(Strain
Energy)
and
equipartition
theorem
reported
in
earlier
work
as


k
T
l
n
(/)
A
a3

h
B
2
t
, where A is the diameter of RBC, a is the minimum spatial
ht2
wavelength that we measure by DPM. The tangential MSD
is measured along the
edge of an RBC where the planar and tangential components can be decoupled (see
Fig.2B inset).
REFERENCES
7
Acknowledgements
We gratefully acknowledge technical discussions with R.R. Dasari and M. Dao, and
helpful suggestions on data analysis from J.P. Mills. This research was supported by the
National Institutes of Health (Spectroscopy Laboratory, P41-RR02594-18), National
Institutes of Health (1-R01-GM076689-01), and the Interdisciplinary Research Group on
Infectious Diseases which is funded by the Singapore-MIT Alliance for Research and
Technology Center.
Competing financial interests
The authors declare no competing financial interest.FIGURE CAPTIONS
Figure 1 RBC topography (a-d) and effective elastic constant map (e-h) for various
stages of the parasite. (a,e), healthy red blood cell. (b, f), ring stage. (c, g), trophozoite.
(d, h), schizont stage. Colorbar indicates thickness in micrometers for a-d, and effective
elastic constant in J/m2. Black arrows indicate the location of parasites, and gray arrow
indicates the location of Hemozoin. (Bright-field and fluorescence micrograph provides
information for locations of parasites and Hemozoin; see Supplemental Figure 1.)
Scalebar is 1.5 m.
Figure 2 Quantitaive membrane fluctuations of membrane and shear modulus. (a),
Histogram of displacement of each stage of parasite infected RBCs. (Inset) Full width
half maximun (FWHM) of fluctuation histogram. (b) Shear modulus of RBC membrane.
(Inset) Schematic of displacements in normal and tangential components.
Figure 3 Full width half maximum (FWHM) of histogram of the membrane
8
fluctuation displacement v.s. temperature for various parasite stage. Groups are the
development stage of P. Falcifarum: healthy RBC (gray), ring-stage (blue), trophozoitestage (green), and schizont (red) The error bar indicate cell to cell variability (N=20 for
all groups). * indicates p-value is less than 0.001, ** less than 0.01
Figure 1
9
Figure
2
1
0
a
0.014
0.012
P(h)
0.01
Healthy
Ring
Trophozoite
Schizont
0.008
0.006
0.004
0.002
0
-200 -150 -100
-50
0
50
h (nm)
100
150
200
b
Modulus (N/m)
100
10
Healthy
Ring
Trophozoite
Figure 3
1
1
Schizont
250
*
Fluctuation (nm)
200
*
150
100
*
*
50
**
*
*
0
*
37°C
23°C
41°C
Supplemental Material
1. Identifying the infected RBCs
In order to identify P. Falciparum infected RBCs, we utilized both bright-field and
fluorescence microscopy as depicted in Fig. S1. Fig. S1a-d show bright-field images of
healthy RBC, ring-stage, trophozoite, and schizont. To distinguish schizont from
trophozoite, we used the Hoest staining and fluorescent microscopy as in Fig. S1 e-f.
1
2
a
b
c
d
e
f
Figure S1. Identifying the infected RBCs by fluorescence. a-d: Bright-field images; e-f:
Hoest stained fluorescent images a: normal b:ring stage; c,e; trophozoite stage and d,f;
schizont stage. Scale bar = 1.5 um
2. Masking the parasite
As P. Falciparum parasites have different refractive index from the RBC hemoglobin
solution, the movement of P. Falciparum could cause artifact in quantifying the motion of
RBC membrane. To minimize this effect, we used the masks for the region where the
parasites are located and excluded those areas from the calculation of the mean squared
displacements. for analyzing the membrane dynamics. The procedure to generate the
masks are illustrated in Fig. S2. First, we indentified the shape and size of RBC using
bright-field microscopy (Fig. S2a) and made mask for the outer shape of the cell (Fig.
S2b). Fluorescent microscopy provide the information about where P. Falciparum
parasites are (Fig. S2c), from which we generate the mask for the parasites. By
subtracting the mask for parasites from the mask for the RBC, we can get the mask for
the parasite-free region (Fig. S2e). Additional smooth filters are used to minimize the
artifact coming from shape edge.
1
3
Figure S2. Masking the parasite in the PDM image. a: bright-field image, b: mask from
a, c: fluorescent image, d: mask from c, e: subtraction d from b, f: applying smooth filter
Movie S3. Movie clip of membrane fluctuations of each stage of P. falciparum
development.
1
4